US5226000A - Method and system for time domain interpolation of digital audio signals - Google Patents
Method and system for time domain interpolation of digital audio signals Download PDFInfo
- Publication number
- US5226000A US5226000A US07/708,912 US70891291A US5226000A US 5226000 A US5226000 A US 5226000A US 70891291 A US70891291 A US 70891291A US 5226000 A US5226000 A US 5226000A
- Authority
- US
- United States
- Prior art keywords
- signal
- digital
- analog
- samples
- time domain
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Lifetime
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/17—Function evaluation by approximation methods, e.g. inter- or extrapolation, smoothing, least mean square method
Definitions
- the present invention relates generally to the field of digital audio systems. More particularly, the present invention relates to a method and system for time domain interpolation of digital audio signals that will allow a digital audio system to resonstruct an analog power audio output signal directly from a digital audio signal.
- Digital audio systems are well known in the prior art.
- CD compact disc
- DAT digital audio tape
- CD compact disc
- DAT digital audio tape
- any digital audio system is to sample and reconstruct an analog audio signal, without noticeable changes to the signal, to recreate authentic sounding music. If, for example, the audio signal is sampled at a recording studio and the digital samples are stored on a CD, then the CD player must retrieve the digital samples and reconstruct the waveform of the audio signal as closely as possible to the waveform of the original analog signal.
- any analog signal can be reconstructed if an infinite number of digital samples are taken of the analog signal.
- the sampling rate of a digital audio system is governed by the Nyquist Theorem that any signal may be sampled and reconstructed provided the sampling rate is at least twice the highest frequency component of the original analog signal.
- An insufficiently high sampling rate tends to create an overlap in the reconstructed signal that gives rise to a special form of distortion known as aliasing.
- aliasing When the sampling rate is too low, the frequency domain images of the reconstructed signal overlap with the baseband and corrupt the higher frequency components of the baseband. Avoidance of aliasing is a primary goal of the sampling process of a digital audio system.
- the human ear is incapable of detecting steady frequencies above 20 KHz, this does not mean that audio signals can be routinely bandlimited to this amount and still achieve high quality audio reproduction.
- studies have indicated that the human ear can perceive sonic effects of transient components of audio signals up to frequencies as high as 100 KHz.
- an audio signal comprised of many transient pieces of high frequency sinusoids is passed through a digital audio system limited to a 20 KHz bandwidth, the transients will be spread out and will lose their transient nature, thereby degrading the quality of the audio reproduction if the digital audio system does not provide for some thpe of correction to the reconstructed signal in the time domain.
- Transients are necessary for professional and high-end audio reproduction because they are important to human hearing in the reconstruction ofd wavefronts that yield the three0dimensional ambience associated with stereophonic signals.
- it is critical that the reproduced music posses this three-dimensional ambience where each individual sound source is perceived as being located on an imaginary sound stage. Indeed, the illusion of a stable three-dimensional sound image is the fundamental feature on which stereo sound is predicated.
- Transients are also important in the resolution of the individual nuances of each of the sound sources. Natural music consists of characteristic noises and momentary silences between notes or overtone oscillations. it is important to prevent sonic blurring of these subtle nuances in the program material. Such details are easily destroyed by audio systems with poor transient response or excessive thermal noise and distortion, with the reproduced music sounding muddy and devoid of fine detail.
- the required bandwidth to pass one cycle of a 15 KHz sinusoidal signal would be 30 KHz, a frequency much higher than the 20 KHz bandwidth limit of current digital audio system.
- a time domain correction could be utilized to more acccurately reconstruct digital audio signals so as to preserve the high frequency transients associated with musical information.
- a frequency domain method of digital audio signal reconstruction should work if the low pass brickwall filter could ideally pass all signals below its threshold or roll-off frequency at unitary gain and reject all signals above its roll-off frequency, and if the distance between the digital sample points is small enough that information is not lost during the sampling process.
- an ideal low pass filter can not be realized. While it is possible to create a low pass brickwall filter that has excellent frequency domain specifications when driven by constant-energy-envelope sinusoids, when this brickwall or taut filter is driven by the transients and impulses of dynamic music material it generates overshoot, ripple and ringing.
- oversampling is used by some prior art digital audio systems to increase the sampling rate to a rate typically four times the original sample rate (e.g., 176 KHz for CD's).
- the basic idea of the prior art oversampling techniques is to implement a digital low pass filter to carry out the function of the analog brickwall smoothing filters, with samples retrived from the digital low pass filter at the higher oversampling rate. This is possible by adding zero magnitude (trivial) samples between each of the original samples to effectively increase the sampling rate of the system, althought the trivial samples add no new information to the signal.
- the problem with current frequency domain oversampling techniques is that the digital fitler, sometimes referred to as a Finite Impulse Response (FIR) filter must meet the same stringent ideal demands as the analog brickwall filter it replaces. Any deviation from an ideal low pass filter will cause corresponding alteration of the output signal.
- Such normalized frequency parameters are too small for the calculations required to derive the associated filter because the numbers do not contain enough significant digits. Without a sufficient number of significant digits in the calculation, these parameters introduce deviation from the desired response. As a result, the frequency domain design method for the digital FIR oversampling filters is unable to accommodate high oversampling rates or a correction to the reconstructed signals in the time domain.
- a further problem in the design of the FIR digital filters in the frequency domain is the arbitrary nature of choosing the appropriate frequency domain parameters. For example, with a given FIR filter order (typically 100 taps), parameters for each of the pass band, transition band, and stop band characteristics must be weighed in the specification of the filter. Without knowing reliable, acceptable figures for these parameters, the designer is effectively guessing at appropriate values for the filter.
- a digital audio system for reconstructing high quality audio signals includes a signal processor means and a direct power output digital-to-analog conversion means.
- the signal processor means reconstructs a digital audio signal by interpolating the digital audio signals in the time domain to allow for the proper recreation of the original audio signal and the direct power output digital-to-analog conversion means generates an analog power audio output signal from the interpolated digital audio signal that can directly drive a speaker to produce the sound waves represented by the audio output signal.
- the signal processor means includes input means for receiving and decoding the digital audio signals comprised of a receiver means for receiving the digital audio signal, phase lock loop means for extracting timing information from the digital audio signal, and decoder means for demodulating the digital audio signal.
- the signal processor means also include a digital processor means for performing a time domain interpolation on the decoded signal to produce an interpolated data signal having an increased sampling rate over the original digital audio signal and a digital volume control means for adjusting the volume of the audio output signal that will drive the speaker by digitally adjusting the interpolated data sample.
- the direct power output digital-to-analog conversion means then converts the volume adjusted interpolated data signal into an analog power audio output signal that may be directly transmitted to the speaker to produce sound.
- the digital processing means is comprised of two parallel signal processors, one for each channel, each signal processor itself being comprised of a pair of interleaved digital signal processors, each performing the calculations for either the even or odd coefficients of a spline-based time domain interpolation.
- the outputs of the signal processors are directly converted to an analog power audio output signal by a series of cascaded co-linear digital-to-analog converters (DACs) immediately coupled to one or more instantaneous current-to-voltage operational amplifier converters whose output is operable connected to the summing junction of the operational amplifier converters to form a single resistive feedback network.
- DACs digital-to-analog converters
- the result is a figital audio system capable of reconstructing hte high frequency and transient characteristcs of the digital audio signals in the time domain and directly converting the reconstructed digital signal into a power analog audio output signal, thereby enabling the reproduction of high-quality musical sound ina professional and high-end digital audio system without the need for further amplification of the audio output signal by an analog amplifier.
- a primary objective of the present invention is to provide a method and system for time domain interpolation of digital audio signals that will allow a digital audio system to more precisely reconstruct an analog power audio output signal from a digital audio signal such that there will be no perceptible difference between the reconstructed signal and the original signal.
- Another objective of the present invention is to minimize time dispersion problems associated with the reconstruction of the digital audio signal by providing a method and system for performing a time domain underpolation to reconstruct the digital audio signal.
- An additional objective of the present invention is to provide a method and system for interpolating digital audio signals that will allow a digital audio system to reproduce an analog power audio output signal directly from a digital audio signal reconstructed in the time domain such that there is no need for further amplification of the analog power audio output signal.
- FIG. 1 is an overall block diagram showing the relationships among the components of a signal processing means in accordance with the present invention.
- FIG. 2 is a graphic representation of a segment of digitized music material represented in the standard AES/EBU digital audio data format.
- FIGS. 3a and 3b are a time and frequency domain representation of a sampled audio signal.
- FIGS. 4a and 4b are a time and frequency domain representation of the sampled audio signal of FIGS. 3a and 3b showing the desired spectrum of the sampled signal.
- FIG. 5 is a time domain representation of the sampled audio signal of FIG. 3a showing the original sampled signal with zero magnitude (trivial) samples added.
- FIG. 6 is a block diagram of an oversampling system.
- FIG. 7 is a frequency domain representation of an ideal low pass filter used to extract the baseband signal during the digital-to-analog conversion.
- FIG. 8 is a signal flow graph of an interpolation system utilizing a time domain interpolation process in accordance with the present invention.
- FIG. 9 is a diagrametric representation of time domain interpolation using the spliine-based algorithm of a preferred embodiment by convolution with zero padded samples.
- FIGS. 10a 10b and 10c are a time chart representation of the time available for the time domain interpolation process utilized in the present invention.
- FIGS. 11a and 11b are a flow chart showing the program flow for a pair of digital signal processors performing the time domain interpolation in accordance with the present invention for a single channel.
- FIG. 12a is a diagrammatic representation of the effective functional blocks of the classic cubic spline-based interpolation method.
- FIG. 12b is a diagrametric representation of the effectve functional blocks of a spine-based filter constructed according to the algorithm shown in FIG. 9.
- the digital audio system includes a signal processor means 10 for porcessing stereo digital audio signals and a direct power output digital-to-analog conversion means 30 for outputting analog power audio output signals to drive a speaker 40.
- FIG. 1 shows only the digital-to-analog conversion means 30 and speaker 40 for one of the channel output signals (the right channel signal).
- the signal processor means 10 and direct power output digital-to-analog conversion means 30 of the present invention are equally applicable to monaural and multi-track audio signals as well.
- the signal processor means 10 includes input means 11 for receiving and decoding the digital audio signals in the form of Input Data from a digital audio signal source.
- the Input Date represents digitized samples of the musical material transmitted in a predetermined stero format (i.e., both left and right channel).
- the Input Data enters a receiver means 12 that receives the Input Data and then synchronizes the Input Data with an internal clock signal via a phase-lock loop means (PLL) 14.
- PLL phase-lock loop means
- a decoder means 16 performs the data demodulation and format decoding of the Input Data to extract the Signal and Timing Data portions of the Input Data.
- the Signal and Timing Data are then processed by a digital processor means 18 that utilizes two digital signal processor means (DSPs) 20 and 22 to perform the time domain interpolation that results in an Interpolated Data Signal.
- the volume of the Interpolated Data Signal generated by the digital processor means 18 is set by a digital volume control means 24 which digitally alters the Interpolated Data Signal to Produce a Volume Adjusted Data Signal.
- the digital signals that comprise the Volume Adjusted Data Signal are then converted into a pair of differential power analog audio Output Signals by the direct power output digital-to-analog conversion means 30.
- the Input Data is provided to the signal processor means 10 by a remotely located professional fiber optic transmitter (not shown) that utilizes graded index, 62.5 micron, glass optical fiber cable and metal/ceramic precision lens professional quality ST® connectors avalable from AT&T, Allentown, PA.
- the receiver means 12 also uses professional quality ST® connectors, thereby creating a high-performance link for receiving the Input Data that is designed to accommodate data rates up to 50 Mbits/second and distances up to 3 km.
- the receiver means 12 further includes a daisy chain optical output connector (not shown) that also uses the professional quality ST® connectors, thereby allowing the digital audio system to ge linked with other similar digital audio system as part of a multi-room audio installation, for example.
- a daisy chain optical output connector (not shown) that also uses the professional quality ST® connectors, thereby allowing the digital audio system to ge linked with other similar digital audio system as part of a multi-room audio installation, for example.
- phase lock loop means 14 is implemented as a frequency lock loop having a crystal controleld oscillator which tracks an internal clock signal to the input clock frequency. It will also be recognized that many other types of phase lock loops or clock recovery schemes could be used to accomplish the same purpose.
- the decoder means 16 demodulates and decodes the Input Data to separate the Input Data into Signal and Timing Data depending upon a preselected format.
- the decoder emans 10 of the signal processor means 10 may ny selectably programmed to received CD Input Data (44.1 KHz), R-DAT Input Data (48 KHz), or Satellite Input Data (32 KHz).
- the sample Format for Input Data received in the standard AES/EBU digital audio format shown in FIG. 2 demonstrates the relationship between the Signal Data and the Timing Data as each are represented in the particular format for the Input Data.
- the decoder means 16 also provides the basic clock and framing signals to the rest of the components of the signal processing means 10.
- the decoder means 16 is implemented using a first programmable gate array that incorporates the necessary digital logic to perform the standard decoding and demodulation operations for AES/EBU digital audio format. It will be recognized that there are also many commercially available decoders for standard digital audio signals which can accomplish this same function.
- AES/EBU digital audio format reference is made to AES Recommended Practice for Digital Audio Engineering--Serial Transmission Format for Linearly Represented Digital Audio Data, ANSI Standar 4.40-1985 which is fuly incorporated herein by reference.
- the digital processor means 18 is comprised of two parallel digital signal processor means 20 and 22, each digital signal processor means consisting of a pair of digital signal processor integrated circuits, oscillators, ROMS, and supporting buffers adn latches.
- digital signal processor means 20 performs the time domain interpolation for the left channel signal and is comprised of a pair of DSPs, DSP0 and DSP1 - Left.
- Digital signal processor means 22 performs the time domain interpolation for the right channel signal and is also comprised of a pair of DSPs, DSP0 and DSP1 - Right.
- DSP0 and DSP1 - Left and DSP0 and DSP1 - Right are comprised of four WE DSP16 chips, also available from AT&T, the operation of which is more fully described in the Data Sheet for the WE DSP16, Oct. 1986, which is fully incorporated by reference herein.
- the digital signal processor means 20 and 22 generate the Interpolated Data as 16 bit digital samples. These 16 bit samples are fed into the digital volume control means 24 and a 20 bit result in the form of the Volume Adjusted Data is produced as a result of the multiplication of the Interpolated Data with a 7-bit left or right channel Digital Volume Control Value.
- the Volume Adjusted Data from the digital volume control 24 is then immediately sent to the direct power output digital-to-analog conversion means 30 for outputing the analog power audio output signals that directly drive the speaker 40.
- the digital volume control means 24 accomplishes a digital volume adjustment by means of a separate hardware multiplication of the Interpolated Data with the Digital Volume Control Value for the respective channel.
- the 7-bit Digital Volume Control Value allows for 128 linear volume control increments to be implemented by the digital volume control means 24.
- the digital volume control means 24 is implemented using a shift and add-type algorithm contained in a second programmable gate array. This second programmable gate array can also handle incidental channel latching and signal connection functions between the signal processor means 10 and the direct power output digital to analog conversion means 30.
- the function of the digital volume control means 24 is accomplished by way of a software routine within the digital signal processing means 20 and 22 that performs the same multiplication of the Interpolated Data with the Digital Volume Control Value.
- the separate hardware digital volume control means 24 can be used to maximize the time available to the digital signal processor means 20 and 22 for performing the time domain interpolation of the present invention, whereas the software digital volume control means 24 can be used to minimize the amount of circuitry external to the digital signal processors 20 and 22 which is required by the present invention.
- the direct power output digital-to-analog conversion means 30 is configured as a differential digital-to-analog converter having a transversal, summed-multiport analog delay line comprised of four DACs 32, one pair of DACs 32 for each rail voltage of the differential signal that together produce an analog current output (I out ) which is then immediately amplified to an analog power audio output signal (V out ) by the respective power current-to-voltage operational amplifier means 34.
- the power current-to-voltage operational amplifier means 34 includes a single resistive feedback means 36 connected between the output terminal and the summing junction input of the power current-to-voltage operational amplifier means 34 for providing a single analog feedback signal within the direct power output digital-to-analog conversion means 30.
- the DACs 32 are four conventional 20-bit co-linear DACs, for example a PCM63P available from Burr-Brown of Arlington, Arizona, or four high quality 20-bit resistor ladder DACs, for example the AD1862 available from Analog Devices of Norwood, Mass.
- the Power current-to-voltage operational amplifier means 34 in a preferred embodiment is comprised of a set of four parallel LN12C operational amplifiers available from National Semiconductor of San Jose, CA., that are driven by series connection from a OPA455BM operational amplifier, also available from Burr-Brown.
- the DACs 32 are connected so as to form a 4 ⁇ delay line by summing their outputs together.
- the first DAC 32 is delayed by 1/4th of the frame time of the frame of interpolated data just calculated by the signal processor 10.
- the second DAC 32 is delayed 1/2 of the frame time
- the third DAC 32 is delayed 3/4th of the frame time
- the fourth DAC 32 is delayed one full frame time.
- the data from the first and third DACs 32 forms the output signal at the positive terminal of the differential Audio Output Signal (V out+ ), while the data from the second and fourth DACs 32 forms the output signal at the negative terminal of the differential Audio Output Signal (V out- ).
- signal processing means 10 and direct power digital-to-analog conversion means 30 are shown in the embodiment described herein as separate from the means for retrieving the Input Data from a recording media (not shown) and the speakers 40 for converting the analog power audio output signals into sound waves, it is also possible to incorporate the signal processing means 10 and direct power digital-to-analog conversion means 30 within a digital audio system that included these components. It is also possible to use the digital audio system of the present invention in the transmission of live music material, for example, a satellite broadcast of a concert. It should be understood that the scope of the present invention includes any combination of the various components that comprise the signal processing means 10 and direct power digital-to-analog conversion means 30, regardless of the ultimate configuration or type of digital audio system with which the present invention may be used.
- T represents the sampling rate.
- the sampled signal has been quantized at the lowest allowable rate specified by the Nyquist therorem.
- the time domain signal of FIG. 5 is identical to that of FIG. 3a with the addition of L-1 samples of zero magnitude between each of the original samples.
- the sampling rate has therefore been increased by a factor of L, but no new information has been added to the sampled signal.
- the frequency domain characteristic is expected to resemble that of the original signal.
- FIG. 6 Performance of the system of FIG. 6 can be analyzed as follows. Assume that the orignal sampling period is equal to T and the interpolated sampling period is T'. If the sampling rate increase is a factor of L then,
- the first block in FIG. 6 carries out the operation of inserting L-1 samples of zero magnitude between each of the original pairs of samples.
- the resultant signal, w(m) is shown in FIG. 5.
- the signal w(m) can be related to the original signal, x(m), as follows:
- this filter would be an iedal lowpass filter with cutoff at 2 ⁇ /T and period T' as shown in FIG. 7.
- the frequency domain signal of FIG. 4b would be obtained. This signal is exactly the signal obtained if the original analog waveform had been sampled at the higher rate, T'.
- the system of FIG. 6 effectively increases the sampling rate by a factor of L.
- the present invention overcomes these deficiencies in the prior art by designing the interpolation filter in the time domain where the elusive optimal frequency domain filter parameters are not required.
- the interpolating filter primarily fulfills a curve fitting function rather than a frequency domain filtering operation and the paramount concern should be the error in the interpolated data.
- Traditional frequency domain filter design does not satisfactorily treat this concern.
- a time domain design procedure for designing the interpolation filter in accordance with the present invention will perform the smoothing of the filter designated h(n) in the system shown in FIG. 6.
- the interpolation of the sampled digital audio signal is performed as a time domain interpolation that involves the generation of a polynomial that passes through Q original sample points along the magnitude curve of the signal as represented in the time domain.
- the signal processing means 10 computes a running set of parameters that fit a curve to the contour of Q original samples, somewhat like pushing a french curve along the sample points and determining the best fit curve for the next set of sample points and then drawing in that curve by filling in the desired number of new samples between each of the original sample points.
- Lagrangian and spline-based interpolations Two possible non-iterative interpolation schemes for providing a digital filter by use of time domain interpolation are Lagrangian and spline-based interpolations.
- the frequency response of any time domain interpolation method must be checked of course, but it can be expected to be good for either Lagrangian or spline-based interpolation, because the original waveform is closely approximated, given that it is bandwidth-limited.
- Linear phase is also a desirable aspect of a time domain digital filter, especially for audio, as it is in effect a constant time delay for all frequencies. There is no actual phase distortion, the signal is simply delayed.
- the Lagrangian interpolation set forth in the previously identified parent application has been formulated as a linear phase FIR filter.
- the spline-based interpolation method of a preferred embodiment of the present invention as described hereinafter has also been designed to have this characteristic.
- Lagrangian and spline-based interpolation algorithms reference is made to Lee W. Johns, R. Dean Riess, Numerical Analysis, Addison-Wesley (1982), pp. 237-247.
- a new set of points and therefore a new polynomial (or possibly the same one) is used for the next interval, and so forth.
- the reconstructed curve produced by this type of Lagrangian interpolation scheme may tend to exhibit oscillatory behavior, especially around the first and last few sample points. This means that although the oscillatory behavior is typicall small in the critical "center" interval [X k , X k-1 ], the curve produced in that center interval is necessarily affected to a certain degree by the error near each end of the group of samples. Specifically, some of the derivatives are poorly estimated at all sample points when the Lagrangian interpolation algorithm is used.
- a spline-based interpolation scheme is used.
- Splines are considered to be good for reconstruction of sampled smooth functions, and especially good at estimating the derivatives at the sample points.
- a spline-based algorithm is generally based on a set of cubic polynomials joined at the sample points. They form a curve which is piecewise continuous, with continuous first and second derivatives throughout, even at the sample points. The third derivative is of course discontinuous in general, being a constant within each interval.
- a spline forms the same curve as that formed by an ideal thin elastic rod held fixed at physical locations defined by ⁇ X i , Y i ⁇ , the sample points. It minimizes the energy of curvature by minimizing the integral of the square of the second derivative.
- a lack of oscillatory behavior is one benefit of using the spline-based algorithm.
- Splines are usually calculated by starting with a discrete set of n points, equally spaced, and solving a matrix equation which determines a set of low-order polynomials, which are usually cubics.
- Each of the n-1 functions interpolates between an adjacent pair out of the n points.
- the process is not computationally complex, but does require some computation time.
- a complete spline does more than may be necessary because it interpolates between all of the points generating a function for each interval, whereas only one function and one sample of that function for the "center" interval is necessary at any given instant to successfuly interpolate a digital audio signal.
- the accuracy of the interpolation is improved by using as many points as possible or practical.
- the accuracy is also best near the center interval [X k , X k+1 ], as more information around the center interval is available.
- the function which interpolates this interval is sampled m times. When that interval is complete, a sample point on one end of the set is dropped (Y 0 or Y n-1 ), and a point is appended to the other end.
- This is a process equivalent to convolution, with zero-padded samples, given that the interpolation can be formulated as a weighted sum. The process is repeated indefinitely, with ⁇ Y i ⁇ in effect shifting across an arbitrarily large set of samples.
- the first n-1 sample points can be used for one spline, to estimate the derivatives or Y n/2-1 , and the last n-1 points for Y n/2 .
- the "next" interval is [Y n/2 , Y n/2+1 ]
- the same n-1 points are used to estimate Y' n/2 each time its derivative is estimated, thereby ensuring continuity of the first derivatives.
- y and y" are length n-1 and n-3 vectors.
- y" is a column vector defined by:
- y is also a column vector defined by:
- Equation 10 For the derivation of a more general case of Equation (10), reference is made to Lee W. Johns, R. Dean Riess, Numerical Analysis, Addison-Wesley (1982), pp. 237-247.
- the last major step in this method is to fit a spline-based polynomial to y k and y k+1 using the estimates for y k' and y k+1 ' set forth in Equation (23).
- Equation (23) there are four unknowns and four equations.
- the unknowns are the four coefficients of:
- Equations 24-29 can be put into a compact matrix form by combining Equations (24) and Equations (25) on the first row, and placing Equation (26) to Equation (29) on the last rows: ##EQU9##
- Equation (31) It is necessary to solve Equation (31) for g to produce the interpolating cubic in the form [t k , t k+1 ]. Because matrix H of Equation (31) is invertible with: ##EQU10## therefore:
- Equation (24) and Equation (33) yields: ##EQU11##
- Equation (34) in matrix results in: ##EQU12##
- the time domain FIR filter coefficients for a preferred embodiment are now found by evaluating Equation (35) for each of the m values of r: 0, 1/m,2/m, . . . , (m-1)/m, and then interleaving the results.
- Equation 35 the values of the coefficients generated by Equation 35 will change depending upon the resolution and smoothing desired, and on the format in which the digitized audio signals are stored. It is also instructive to note that the number of coefficients necessary for a given number of samples Q and sampling rate increase L is equal to,
- FIG. 8 A signal flow graph of the basic interpolation system for such a time domain interpolation system is shown in FIG. 8. Such a network is equivalent to a traditional FIR filter design.
- the spline-based time domain interpolation can be implemented as an FIR topology.
- the time domain interpolation is calculated for a 3rd order polynomial to fit the center interval of the contour of twelve (Q) original samples at a time by generating 15 new samples between each of the original twelve samples.
- the signal processing means 10 of the present invention is able to perform a time domain interpolation of a segment of the audio signal and can achieve the oversampling rates necessary to enable the high-quality reproduction of musical sound in a professional or high-end digital audio system.
- the present invention utilizes a unique approach to the design and implementation of the audio signal reconstruction process. This approach emphasizes certain design constraints that are different from the prior art digital audio systems.
- a goal of the present invention is to guarantee optimal transient response.
- Traditional frequency domain oversampling cannot incorporate time-domain transient parameters into its design.
- the response of a time domain oversampling system can be constrained to pass all original sample points guaranteeing that no significant overshoot or widening of transients will occur.
- each of the digital signal processors, DSP0 and DSP1 contains an arithmetic unit with a 16 bit ⁇ 16 bit parallel multiplier that generates a full 32 bit product in 55 ns. The product can be accumulated with one of two 36 bit accumulators.
- Each DSP operates at a clock frequency of 36 MHz. Consequently, the present invention utilizes a parallel pipeline architecture on each channel to operate two DSPs, DSP0 and DSP1, in parallel to achieve the required 72 MIPS processor speed.
- the pair of DSPs operate in tandem with DSP0 used to calculate the even sample vaues and DSP1 used to calculate the odd sample values, thereby effectively doubling the amount of time available for each processor to complete the necessary calculations before moving to the next value.
- the time available to compute each output for the Interpolated Data output is T avail .
- the processor clock speed of each DSP is only 36MHz, the calculation of the coefficients for the desired equation cannot be completed in T avail . If however, the calculations were performed for every other output, then, as shown in FIG. 10b, the effective T avail is doubled.
- the present invention achieves both the speed and the precision necessary to perform the required calculations and drive the high length filter that comprises the digital conversion means 30 and 32.
- Each DSP is loaded with the identical interpolation program, with the calculations for the relevant coefficients for the even and odd outputs split between the two DSP's such that DSP0 calculates the even indexed coefficients and DSP1 calculates the odd-indexed coefficients.
- the result is a time domain interpolation that can produce the interpolated samples at twice the rate of a single DSP. It will be apparent that this method of interleaving the DSP's to calculate the outputs for each new set of samples could be extended to more than just a pair of DSP's operating in tandem. In general, a time domain interpolation simulating an FIR filter of any length could be implemented by this technique. If the desired FIR filter was of length N, and a single DSP could compute M terms between input samples, then N/M DSPs would be required in the parallel method described.
- each DSP is initialized and memory is allocated for the working variables used by the processors.
- the program will assume zero values for the first twelve (Q) samples so that the samples at the beginning of the segment of digitized music material can also be interpolated using the same procedures as the remaining samples.
- a pointer is set to a circular buffer containing the predetermined values for the particular coefficients of Equation (38) that will be used in performing the interpolation.
- the dithering process takes the place of a normal rounding processing during the calculation, but produces a statistically better time average for estimating the interpolated value of the signal.
- the final value of the first interpolated sample of the Interpolated Data output is rounded to 16-bits to be transferred to the digital volume control means 24.
- the first interpolated sample output is transferred to the digital volume control means 24 and onto the direct power output digital to analog conversion means 30.
- DSP 1 computes the values for the second interpolated sample output in accordance with the predetermined coefficients pointed to by the pointer for DSP 1.
- the second interpolated sample is dithered at Dither 116 and rounded at Round 118 to be output at Output 120 as the final value of the second interpolated sample of the Interpolated Data output.
- DSP1 has an adequate amount of time in which to perform the necessary calculations. This process is repeated until all sixteen interpolated sample outputs have been generated, which each of the DSPs having sufficient time to perform the calculations necessary to achieve an effective 16 ⁇ oversampling rate.
- the Interpolated Data output by the pair of DSPs, DSP0 and DSP1, for each channel implements the 192 tap Finite Impulse Response (FIR) structure in accordance with equation (38).
- FIR Finite Impulse Response
- An additional advantage of using the time domain interpolation of the embodiment of the present invention as described therein is that the dithering of the interpolated values is accomplished on a 16-bit sample value and the additional increase in bit resolution resulting from the digital volume control means 24 and the direct power digital to analog conversion means 30 is performed after the dithering process is accomplished.
- This method of performing dithering on a digital sample that has the same bit-resolution as the original digital sample contained in the digital audio signal results in a more accurate reproduction of the original audio signal.
- dithering is performed on the interpolated samples at a bit resolution higher than the bit resolution of the original digital sample (e.g., dithering at 18-bit or 20-bit resolution).
- these digital audio systems introduce dithering error into the reconstruction of the digital audio signal by assuming a higher degree of resolution than is actually present in the original digital sample.
- the present invention achieves the same or better bit resolution without incurring suc dithering error by using the digital volume control and delay line digital to analog conversion techniques to increase the bit resolution of the Interpolated Data after the interpolated samples have been dithered.
Abstract
Description
T=T/l (Eq. 1)
w(m)={x(m/L) m=0, +/-L, +/-2L } (Eq. 2)
W(Z)=X(Z.sup.L) (Eq. 4)
(n.sub.even) (X,Y)0 . . . (X,Y).sub.n-1 (Eq. 6)
(X.sub.k, X.sub.k-1), where k=(n/2)-1 (Eq. 7)
Ay"=(6/h.sup.2)By (Eq. 10)
y"=(y.sub.1 ", y.sub.2 ", . . . , y.sub.n-3 ").sup.t (Eq. 11)
y=(y.sub.0, y.sub.1, . . . y.sub.n-2).sup.t (Eq. 12)
y"=(6/h.sup.2)A.sup.-1 By (Eq. 13)
S.sub.i (t)=(y.sub.i "/6h)(t.sub.i+1 -t).sup.3 +(y.sub.i+1 "/6h)(t-t.sub.i).sup.3 +((y.sub.i+1 /h)-(hy.sub.i+1 "/6))(t-t.sub.i)+((y.sub.i /h)-(hy.sub.i "/6))(t.sub.i+1 -t) (Eq. 14)
S'.sub.k (r)=y"h(r-r.sup.2 /2-1/3l)+y.sub.k+1 "h(r.sup.2 /2-1/3)-(1/h)y.sub.k +(1/h)y.sub.k+1 (Eq. 15)
y.sub.k '≈S.sub.k '(0)=-(h/3)y.sub.k "-(h/6)y.sub.k+1 "-(1/h)y.sub.k +(1/h)y.sub.k+1 (Eq. 16)
y.sub.k '≈S.sub.k '(0)=S.sup.(1)T y"+S.sup.(2)T y (Eq. 18)
S.sup.(1)T y"=S.sub.0.sup.(1) y"+S.sub.1 '=S.sub.2.sup.(1) y.sub.2 "=. . . +S.sub.n-4.sup.(1) y.sub.n-3 " (Eq. 19)
y.sub.k '≈((6/.sup.2)S.sup.(1)T A.sup.-1 B+S.sup.(2)T y=d.sup.T y (Eq. 20)
d=((6/h.sup.2)S.sup.(1)T A.sup.-1 B +S.sup.(2)T).sup.T (Eq. 21)
y.sub.k '≈d.sup.(1)T y and
y.sub.k+1 '≈d.sup.(2)T y (Eq. 23)
P.sub.k (t)=a(t-t.sub.k).sup.3 +b(t-t.sub.k).sup.2 +c(t-t.sub.k)+d (Eq. 24)
P.sub.k '(t)=3a(t-t.sub.k).sup.2 +b(t-t.sub.k)+c (Eq. 25)
e.sub.k+1.sup.T y=P.sub.k (t.sub.k+1)=ah.sup.3 +bh.sup.2 +ch+d (Eq. 27)
d.sup.(1)T y=P.sub.k '(t.sub.k)=c (Eq. 28)
d.sup.(2)T Y=P.sub.k '(t.sub.k+1)=ah.sup.2 +2bh +c (Eq. 29)
g=H.sup.-1 Dy (Eq. 33)
N=Q*L (Eq. 36)
Claims (32)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US07/708,912 US5226000A (en) | 1988-11-08 | 1991-05-31 | Method and system for time domain interpolation of digital audio signals |
Applications Claiming Priority (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US26883088A | 1988-11-08 | 1988-11-08 | |
US07/597,512 US5075880A (en) | 1988-11-08 | 1990-10-12 | Method and apparatus for time domain interpolation of digital audio signals |
US07/708,912 US5226000A (en) | 1988-11-08 | 1991-05-31 | Method and system for time domain interpolation of digital audio signals |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US07/597,512 Continuation-In-Part US5075880A (en) | 1988-11-08 | 1990-10-12 | Method and apparatus for time domain interpolation of digital audio signals |
Publications (1)
Publication Number | Publication Date |
---|---|
US5226000A true US5226000A (en) | 1993-07-06 |
Family
ID=27402114
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US07/708,912 Expired - Lifetime US5226000A (en) | 1988-11-08 | 1991-05-31 | Method and system for time domain interpolation of digital audio signals |
Country Status (1)
Country | Link |
---|---|
US (1) | US5226000A (en) |
Cited By (39)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5587711A (en) * | 1994-09-30 | 1996-12-24 | Apple Computer, Inc. | Method and system for reconstructing quantized signals |
US5623402A (en) * | 1994-02-10 | 1997-04-22 | Schenck Pegasus Corporation | Multi-channel inverse control using adaptive finite impulse response filters |
US5719797A (en) * | 1995-12-15 | 1998-02-17 | The United States Of America As Represented By The Secretary Of The Army | Simulator for smart munitions testing |
US5787259A (en) * | 1996-03-29 | 1998-07-28 | Microsoft Corporation | Digital interconnects of a PC with consumer electronics devices |
US5859787A (en) * | 1995-11-09 | 1999-01-12 | Chromatic Research, Inc. | Arbitrary-ratio sampling rate converter using approximation by segmented polynomial functions |
US5903480A (en) * | 1997-09-29 | 1999-05-11 | Neomagic | Division-free phase-shift for digital-audio special effects |
US6483451B1 (en) * | 1999-07-09 | 2002-11-19 | Yasue Sakai | Sampling function waveform data generating device |
US20020183874A1 (en) * | 2001-04-30 | 2002-12-05 | The Government Of The United States Of America As Represented By The Secretary Of The Department Of | Auscultatory training system |
US6708145B1 (en) * | 1999-01-27 | 2004-03-16 | Coding Technologies Sweden Ab | Enhancing perceptual performance of sbr and related hfr coding methods by adaptive noise-floor addition and noise substitution limiting |
US20040147229A1 (en) * | 2001-04-10 | 2004-07-29 | Mcgrath David S. | High frequency signal construction method and apparatus |
US20080228292A1 (en) * | 2007-03-12 | 2008-09-18 | M/A-Com, Inc., | System and method for pre-distorting a device input |
US7610553B1 (en) * | 2003-04-05 | 2009-10-27 | Apple Inc. | Method and apparatus for reducing data events that represent a user's interaction with a control interface |
US9484946B2 (en) * | 2014-08-25 | 2016-11-01 | Nxp B.V. | Digital-to-analog converter (DAC), method for operating a DAC and transceiver circuit |
WO2017222559A1 (en) * | 2016-06-24 | 2017-12-28 | Cirrus Logic International Semiconductor Ltd. | Digital volume control |
US9875750B2 (en) | 2016-06-24 | 2018-01-23 | Cirrus Logic, Inc. | Digital volume control |
WO2021074603A1 (en) * | 2019-10-13 | 2021-04-22 | Ultraleap Limited | Reducing harmonic distortion by dithering |
US11098951B2 (en) | 2018-09-09 | 2021-08-24 | Ultrahaptics Ip Ltd | Ultrasonic-assisted liquid manipulation |
US11169610B2 (en) | 2019-11-08 | 2021-11-09 | Ultraleap Limited | Tracking techniques in haptic systems |
US11189140B2 (en) | 2016-01-05 | 2021-11-30 | Ultrahaptics Ip Ltd | Calibration and detection techniques in haptic systems |
US11204644B2 (en) | 2014-09-09 | 2021-12-21 | Ultrahaptics Ip Ltd | Method and apparatus for modulating haptic feedback |
US11276281B2 (en) | 2015-02-20 | 2022-03-15 | Ultrahaptics Ip Ltd | Algorithm improvements in a haptic system |
US11307664B2 (en) | 2016-08-03 | 2022-04-19 | Ultrahaptics Ip Ltd | Three-dimensional perceptions in haptic systems |
US11341952B2 (en) | 2019-08-06 | 2022-05-24 | Insoundz, Ltd. | System and method for generating audio featuring spatial representations of sound sources |
US11360546B2 (en) | 2017-12-22 | 2022-06-14 | Ultrahaptics Ip Ltd | Tracking in haptic systems |
US11378997B2 (en) | 2018-10-12 | 2022-07-05 | Ultrahaptics Ip Ltd | Variable phase and frequency pulse-width modulation technique |
US11483654B2 (en) * | 2020-02-10 | 2022-10-25 | Cirrus Logic, Inc. | Driver circuitry |
US11531395B2 (en) | 2017-11-26 | 2022-12-20 | Ultrahaptics Ip Ltd | Haptic effects from focused acoustic fields |
US11529650B2 (en) | 2018-05-02 | 2022-12-20 | Ultrahaptics Ip Ltd | Blocking plate structure for improved acoustic transmission efficiency |
US11543507B2 (en) | 2013-05-08 | 2023-01-03 | Ultrahaptics Ip Ltd | Method and apparatus for producing an acoustic field |
US11553295B2 (en) | 2019-10-13 | 2023-01-10 | Ultraleap Limited | Dynamic capping with virtual microphones |
US11550395B2 (en) | 2019-01-04 | 2023-01-10 | Ultrahaptics Ip Ltd | Mid-air haptic textures |
US11550432B2 (en) | 2015-02-20 | 2023-01-10 | Ultrahaptics Ip Ltd | Perceptions in a haptic system |
US11704983B2 (en) | 2017-12-22 | 2023-07-18 | Ultrahaptics Ip Ltd | Minimizing unwanted responses in haptic systems |
US11715453B2 (en) | 2019-12-25 | 2023-08-01 | Ultraleap Limited | Acoustic transducer structures |
US11727790B2 (en) | 2015-07-16 | 2023-08-15 | Ultrahaptics Ip Ltd | Calibration techniques in haptic systems |
US11816267B2 (en) | 2020-06-23 | 2023-11-14 | Ultraleap Limited | Features of airborne ultrasonic fields |
US11842517B2 (en) | 2019-04-12 | 2023-12-12 | Ultrahaptics Ip Ltd | Using iterative 3D-model fitting for domain adaptation of a hand-pose-estimation neural network |
US11886639B2 (en) | 2020-09-17 | 2024-01-30 | Ultraleap Limited | Ultrahapticons |
US11955109B2 (en) | 2016-12-13 | 2024-04-09 | Ultrahaptics Ip Ltd | Driving techniques for phased-array systems |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4076958A (en) * | 1976-09-13 | 1978-02-28 | E-Systems, Inc. | Signal synthesizer spectrum contour scaler |
US4408094A (en) * | 1979-10-16 | 1983-10-04 | Nippon Electric Co., Ltd. | Output circuit |
US4591832A (en) * | 1984-07-18 | 1986-05-27 | Rca Corporation | Digital-to-analog conversion system as for use in a digital TV receiver |
US5046107A (en) * | 1986-09-30 | 1991-09-03 | Yamaha Corporation | Input level adjusting circuit |
US5075880A (en) * | 1988-11-08 | 1991-12-24 | Wadia Digital Corporation | Method and apparatus for time domain interpolation of digital audio signals |
-
1991
- 1991-05-31 US US07/708,912 patent/US5226000A/en not_active Expired - Lifetime
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US4076958A (en) * | 1976-09-13 | 1978-02-28 | E-Systems, Inc. | Signal synthesizer spectrum contour scaler |
US4408094A (en) * | 1979-10-16 | 1983-10-04 | Nippon Electric Co., Ltd. | Output circuit |
US4591832A (en) * | 1984-07-18 | 1986-05-27 | Rca Corporation | Digital-to-analog conversion system as for use in a digital TV receiver |
US5046107A (en) * | 1986-09-30 | 1991-09-03 | Yamaha Corporation | Input level adjusting circuit |
US5075880A (en) * | 1988-11-08 | 1991-12-24 | Wadia Digital Corporation | Method and apparatus for time domain interpolation of digital audio signals |
Non-Patent Citations (6)
Title |
---|
Analog Devices, Inc., "Analog-Digital Conversion Handbook", 1986, p. 237, Prentice-Hall, Englewood Cliffs, N.J. 07632. |
Analog Devices, Inc., Analog Digital Conversion Handbook , 1986, p. 237, Prentice Hall, Englewood Cliffs, N.J. 07632. * |
J. Byerly & M. Vander Kooi, "National Semiconductor Handbook", LM380 Power Audio Amplifier, Dec. 1972, pp. AN69-1 to AN69-7. |
J. Byerly & M. Vander Kooi, National Semiconductor Handbook , LM380 Power Audio Amplifier, Dec. 1972, pp. AN69 1 to AN69 7. * |
Walter G. Jung, "IC Op-Amp Cookbook", 1980, pp. 315-319, Howard W. Sams & Co., Inc., 4300 W. 62nd St., Indianapolis, Ind. 46268. |
Walter G. Jung, IC Op Amp Cookbook , 1980, pp. 315 319, Howard W. Sams & Co., Inc., 4300 W. 62nd St., Indianapolis, Ind. 46268. * |
Cited By (68)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5623402A (en) * | 1994-02-10 | 1997-04-22 | Schenck Pegasus Corporation | Multi-channel inverse control using adaptive finite impulse response filters |
US5587711A (en) * | 1994-09-30 | 1996-12-24 | Apple Computer, Inc. | Method and system for reconstructing quantized signals |
US5859787A (en) * | 1995-11-09 | 1999-01-12 | Chromatic Research, Inc. | Arbitrary-ratio sampling rate converter using approximation by segmented polynomial functions |
US5719797A (en) * | 1995-12-15 | 1998-02-17 | The United States Of America As Represented By The Secretary Of The Army | Simulator for smart munitions testing |
US5787259A (en) * | 1996-03-29 | 1998-07-28 | Microsoft Corporation | Digital interconnects of a PC with consumer electronics devices |
US5903480A (en) * | 1997-09-29 | 1999-05-11 | Neomagic | Division-free phase-shift for digital-audio special effects |
USRE43189E1 (en) * | 1999-01-27 | 2012-02-14 | Dolby International Ab | Enhancing perceptual performance of SBR and related HFR coding methods by adaptive noise-floor addition and noise substitution limiting |
US8255233B2 (en) | 1999-01-27 | 2012-08-28 | Dolby International Ab | Enhancing perceptual performance of SBR and related HFR coding methods by adaptive noise-floor addition and noise substitution limiting |
US6708145B1 (en) * | 1999-01-27 | 2004-03-16 | Coding Technologies Sweden Ab | Enhancing perceptual performance of sbr and related hfr coding methods by adaptive noise-floor addition and noise substitution limiting |
US9245533B2 (en) | 1999-01-27 | 2016-01-26 | Dolby International Ab | Enhancing performance of spectral band replication and related high frequency reconstruction coding |
US8935156B2 (en) | 1999-01-27 | 2015-01-13 | Dolby International Ab | Enhancing performance of spectral band replication and related high frequency reconstruction coding |
US8738369B2 (en) | 1999-01-27 | 2014-05-27 | Dolby International Ab | Enhancing performance of spectral band replication and related high frequency reconstruction coding |
US8543385B2 (en) | 1999-01-27 | 2013-09-24 | Dolby International Ab | Enhancing perceptual performance of SBR and related HFR coding methods by adaptive noise-floor addition and noise substitution limiting |
US20090315748A1 (en) * | 1999-01-27 | 2009-12-24 | Liljeryd Lars G | Enhancing Perceptual Performance of SBR and Related HFR Coding Methods by Adaptive Noise-Floor Addition and Noise Substitution Limiting |
US20090319280A1 (en) * | 1999-01-27 | 2009-12-24 | Liljeryd Lars G | Enhancing Perceptual Performance of SBR and Related HFR Coding Methods by Adaptive Noise-Floor Addition and Noise Substitution Limiting |
US20090319259A1 (en) * | 1999-01-27 | 2009-12-24 | Liljeryd Lars G | Enhancing Perceptual Performance of SBR and Related HFR Coding Methods by Adaptive Noise-Floor Addition and Noise Substitution Limiting |
US8036882B2 (en) | 1999-01-27 | 2011-10-11 | Coding Technologies Sweden Ab | Enhancing perceptual performance of SBR and related HFR coding methods by adaptive noise-floor addition and noise substitution limiting |
US8036880B2 (en) | 1999-01-27 | 2011-10-11 | Coding Technologies Sweden Ab | Enhancing perceptual performance of SBR and related HFR coding methods by adaptive noise-floor addition and noise substitution limiting |
US8036881B2 (en) | 1999-01-27 | 2011-10-11 | Coding Technologies Sweden Ab | Enhancing perceptual performance of SBR and related HFR coding methods by adaptive noise-floor addition and noise substitution limiting |
US6483451B1 (en) * | 1999-07-09 | 2002-11-19 | Yasue Sakai | Sampling function waveform data generating device |
US7685218B2 (en) | 2001-04-10 | 2010-03-23 | Dolby Laboratories Licensing Corporation | High frequency signal construction method and apparatus |
US20040147229A1 (en) * | 2001-04-10 | 2004-07-29 | Mcgrath David S. | High frequency signal construction method and apparatus |
US20020183874A1 (en) * | 2001-04-30 | 2002-12-05 | The Government Of The United States Of America As Represented By The Secretary Of The Department Of | Auscultatory training system |
US7209796B2 (en) | 2001-04-30 | 2007-04-24 | The United States Of America As Represented By The Secretary Of The Department Of Health And Human Services, Centers For Disease Control And Prevention | Auscultatory training system |
US7610553B1 (en) * | 2003-04-05 | 2009-10-27 | Apple Inc. | Method and apparatus for reducing data events that represent a user's interaction with a control interface |
US8050783B2 (en) * | 2007-03-12 | 2011-11-01 | Pine Valley Investments, Inc. | System and method for pre-distorting a device input |
US20080228292A1 (en) * | 2007-03-12 | 2008-09-18 | M/A-Com, Inc., | System and method for pre-distorting a device input |
US11543507B2 (en) | 2013-05-08 | 2023-01-03 | Ultrahaptics Ip Ltd | Method and apparatus for producing an acoustic field |
US11624815B1 (en) | 2013-05-08 | 2023-04-11 | Ultrahaptics Ip Ltd | Method and apparatus for producing an acoustic field |
US9484946B2 (en) * | 2014-08-25 | 2016-11-01 | Nxp B.V. | Digital-to-analog converter (DAC), method for operating a DAC and transceiver circuit |
US11656686B2 (en) | 2014-09-09 | 2023-05-23 | Ultrahaptics Ip Ltd | Method and apparatus for modulating haptic feedback |
US11768540B2 (en) | 2014-09-09 | 2023-09-26 | Ultrahaptics Ip Ltd | Method and apparatus for modulating haptic feedback |
US11204644B2 (en) | 2014-09-09 | 2021-12-21 | Ultrahaptics Ip Ltd | Method and apparatus for modulating haptic feedback |
US11276281B2 (en) | 2015-02-20 | 2022-03-15 | Ultrahaptics Ip Ltd | Algorithm improvements in a haptic system |
US11550432B2 (en) | 2015-02-20 | 2023-01-10 | Ultrahaptics Ip Ltd | Perceptions in a haptic system |
US11830351B2 (en) | 2015-02-20 | 2023-11-28 | Ultrahaptics Ip Ltd | Algorithm improvements in a haptic system |
US11727790B2 (en) | 2015-07-16 | 2023-08-15 | Ultrahaptics Ip Ltd | Calibration techniques in haptic systems |
US11189140B2 (en) | 2016-01-05 | 2021-11-30 | Ultrahaptics Ip Ltd | Calibration and detection techniques in haptic systems |
CN109478892A (en) * | 2016-06-24 | 2019-03-15 | 思睿逻辑国际半导体有限公司 | Digital volume control |
US9875750B2 (en) | 2016-06-24 | 2018-01-23 | Cirrus Logic, Inc. | Digital volume control |
WO2017222559A1 (en) * | 2016-06-24 | 2017-12-28 | Cirrus Logic International Semiconductor Ltd. | Digital volume control |
GB2562819B (en) * | 2016-06-24 | 2021-07-07 | Cirrus Logic Int Semiconductor Ltd | Digital volume control |
GB2562819A (en) * | 2016-06-24 | 2018-11-28 | Cirrus Logic Int Semiconductor Ltd | Digital volume control |
US11307664B2 (en) | 2016-08-03 | 2022-04-19 | Ultrahaptics Ip Ltd | Three-dimensional perceptions in haptic systems |
US11714492B2 (en) | 2016-08-03 | 2023-08-01 | Ultrahaptics Ip Ltd | Three-dimensional perceptions in haptic systems |
US11955109B2 (en) | 2016-12-13 | 2024-04-09 | Ultrahaptics Ip Ltd | Driving techniques for phased-array systems |
US11531395B2 (en) | 2017-11-26 | 2022-12-20 | Ultrahaptics Ip Ltd | Haptic effects from focused acoustic fields |
US11921928B2 (en) | 2017-11-26 | 2024-03-05 | Ultrahaptics Ip Ltd | Haptic effects from focused acoustic fields |
US11360546B2 (en) | 2017-12-22 | 2022-06-14 | Ultrahaptics Ip Ltd | Tracking in haptic systems |
US11704983B2 (en) | 2017-12-22 | 2023-07-18 | Ultrahaptics Ip Ltd | Minimizing unwanted responses in haptic systems |
US11529650B2 (en) | 2018-05-02 | 2022-12-20 | Ultrahaptics Ip Ltd | Blocking plate structure for improved acoustic transmission efficiency |
US11883847B2 (en) | 2018-05-02 | 2024-01-30 | Ultraleap Limited | Blocking plate structure for improved acoustic transmission efficiency |
US11098951B2 (en) | 2018-09-09 | 2021-08-24 | Ultrahaptics Ip Ltd | Ultrasonic-assisted liquid manipulation |
US11740018B2 (en) | 2018-09-09 | 2023-08-29 | Ultrahaptics Ip Ltd | Ultrasonic-assisted liquid manipulation |
US11378997B2 (en) | 2018-10-12 | 2022-07-05 | Ultrahaptics Ip Ltd | Variable phase and frequency pulse-width modulation technique |
US11550395B2 (en) | 2019-01-04 | 2023-01-10 | Ultrahaptics Ip Ltd | Mid-air haptic textures |
US11842517B2 (en) | 2019-04-12 | 2023-12-12 | Ultrahaptics Ip Ltd | Using iterative 3D-model fitting for domain adaptation of a hand-pose-estimation neural network |
US11881206B2 (en) | 2019-08-06 | 2024-01-23 | Insoundz Ltd. | System and method for generating audio featuring spatial representations of sound sources |
US11341952B2 (en) | 2019-08-06 | 2022-05-24 | Insoundz, Ltd. | System and method for generating audio featuring spatial representations of sound sources |
US11742870B2 (en) | 2019-10-13 | 2023-08-29 | Ultraleap Limited | Reducing harmonic distortion by dithering |
US11553295B2 (en) | 2019-10-13 | 2023-01-10 | Ultraleap Limited | Dynamic capping with virtual microphones |
US11374586B2 (en) | 2019-10-13 | 2022-06-28 | Ultraleap Limited | Reducing harmonic distortion by dithering |
WO2021074603A1 (en) * | 2019-10-13 | 2021-04-22 | Ultraleap Limited | Reducing harmonic distortion by dithering |
US11169610B2 (en) | 2019-11-08 | 2021-11-09 | Ultraleap Limited | Tracking techniques in haptic systems |
US11715453B2 (en) | 2019-12-25 | 2023-08-01 | Ultraleap Limited | Acoustic transducer structures |
US11483654B2 (en) * | 2020-02-10 | 2022-10-25 | Cirrus Logic, Inc. | Driver circuitry |
US11816267B2 (en) | 2020-06-23 | 2023-11-14 | Ultraleap Limited | Features of airborne ultrasonic fields |
US11886639B2 (en) | 2020-09-17 | 2024-01-30 | Ultraleap Limited | Ultrahapticons |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US5226000A (en) | Method and system for time domain interpolation of digital audio signals | |
US5075880A (en) | Method and apparatus for time domain interpolation of digital audio signals | |
BR9715316B1 (en) | btsc encoder. | |
EP1639703B1 (en) | Rational sample rate conversion | |
US7369906B2 (en) | Digital audio signal processing | |
JPH06232755A (en) | Signal processing system and processing method | |
US5267095A (en) | Digital reconstructing of harmonics to extend band of frequency response | |
US5497154A (en) | Dither generating apparatus | |
US4044306A (en) | Digital converter from pulse code modulation to continuous variable slope delta modulation | |
JPS5920204B2 (en) | Adaptive delta modulation system | |
US5461604A (en) | PCM digital audio signal playback apparatus | |
US4035724A (en) | Digital converter from continuous variable slope delta modulation to pulse code modulation | |
US5594443A (en) | D/A converter noise reduction system | |
US6531969B2 (en) | Resampling system and apparatus | |
EP0492578B1 (en) | Digital filter | |
US5657261A (en) | Interpolation of digital signals using signal sample replication | |
US5387910A (en) | Signal processor | |
JPH0991887A (en) | Digital signal processing method and device | |
RU2130233C1 (en) | Transmission device which recovers lost signal samples | |
US5774478A (en) | Interpolation circuit for interpolating error data block generated in Σ modulated data stream | |
US6772022B1 (en) | Methods and apparatus for providing sample rate conversion between CD and DAT | |
CA2053133C (en) | Method for coding and decoding a sampled analog signal having a repetitive nature and a device for coding and decoding by said method | |
JPH0732343B2 (en) | Asynchronous sampling frequency conversion method | |
JP3312538B2 (en) | Sound signal processing device | |
JPS5898793A (en) | Voice synthesizer |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: WADIA DIGITAL CORPORATION A CORPORATION OF MN, WI Free format text: ASSIGNMENT OF ASSIGNORS INTEREST.;ASSIGNORS:MOSES, ROBERT W.;MILLER, RAYMOND S.;HUSTIG, CHARLES H.;AND OTHERS;REEL/FRAME:005819/0675;SIGNING DATES FROM 19910717 TO 19910802 |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |
|
CC | Certificate of correction | ||
FPAY | Fee payment |
Year of fee payment: 4 |
|
FPAY | Fee payment |
Year of fee payment: 8 |
|
AS | Assignment |
Owner name: SHARED VENTURES, INC., MINNESOTA Free format text: FORECLOSURE;ASSIGNOR:WADIA DIGITAL CORPORATION;REEL/FRAME:011575/0432 Effective date: 20000912 |
|
REMI | Maintenance fee reminder mailed | ||
REMI | Maintenance fee reminder mailed | ||
FPAY | Fee payment |
Year of fee payment: 12 |
|
SULP | Surcharge for late payment |
Year of fee payment: 11 |
|
AS | Assignment |
Owner name: TRUEWAVE, L.L.C., MICHIGAN Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:SHARED VENTURES, INC.;REEL/FRAME:017344/0043 Effective date: 20060131 |